Every week day 10:45-12:35
Oxley Hall 218 (Computational Linguistics Lab)
Instructor: Marie-Catherine de Marneffe
Office hours: Thursday 1:00 - 2:00 (Oxley 203) or by appointment

In current linguistic research, it is often necessary to deal with lots of data (corpus or experimental data). This course offers practical training in standard computational tools for tackling different kinds of data for linguistic research. Students will learn computational techniques to access, search and format linguistic datasets, including text corpora, speech and audio, structured representations and experimental measurements. The course will also cover data exploration and visualization.

No prerequisite in programming is required: the course will cover introductory scripting in Python, R and Praat. The course is designed to be hands-on, and students will have the opportunity to work on the problem sets during the class sessions.

1. Students will gain hands-on experience gathering, formatting, and manipulating data.
2. Students will learn to use corpus, field, and experimental data, as well as to combine data from multiple sources.
3. Students will learn to work with existing computational tools.
4. At the end of the course, students will be able to process massive amounts of linguistic data.

The course is designed to stand alone, but also to provide an introduction to the graduate Computational Linguistics sequence. It is not a prerequisite for the Computational Linguistics courses, but is helpful for students who lack any prior experience with computational tools.

Accessing and navigating corpora

Linguistic data manipulation and visualization

Automatic processing of structured linguistic representations

R scripting

Praat scripting

The schedule page and this one serve as the syllabus for this course.

Unit 1: Basic data manipulations

• Introduction and motivation
  - Case study: Do women talk more than men?
  - How to use dialogue corpora to test a hypothesis
• Basic Unix environment
  - How to access and navigate data directories
• A computer language to deal with human language: Introductory Python
  - Basic file IO
  - Decision-making: logic, comparatives, conditionals

Unit 2: Reading text and counting words

• Case study: Investigating Zipf's law
• Counting instances of a word in a file
• Dictionaries
• Counting all words/bigrams

Unit 3: R and Praat scripting

• Case study: Automatically extracting measurements to make vowel space plots, a glimpse into OhioSpeaks
• The R language:
  - Variables, control statements and data structures in R
  - Data exploration: the dative alternation, do children differ from adults?
• The Praat language:
  - Variables, control statements and data structures in Praat
  - Manipulation of audio data

Unit 4: Dealing with linguistic structured representations

• Case study: Which verbs allow the dative alternation?
• Field-structured: CSV, space-delimited
• Tree-structured and NLTK: Penn TreeBank parses
• Internationalization, non-standard character sets: How to deal with Arabic, Chinese, Hindi or Cyrillic?

There will be three assignments to turn in. Each one will require students to write a short program to perform some analysis of a dataset (for instance, assignment 1 is to write a Python program measuring utterance lengths by men and women in a section of the Fisher corpus). Students will work on the assignments both in class and at home, and will be encouraged to work collaboratively in small groups, but everyone has to turn in their own assignment.

Periodically there will be day to day short assignments strongly recommended, but that do not have to be turned in.

There will however be three short answers that need to be turned in. The goal of these short answers is to make sure that you are staying caught up with the material.

The course has been designed to be hands-on and attendance is therefore strongly recommended. The meetings happen in person. (Online accommodation is possible, please contact me as soon as possible if you need it).

A tentative schedule for the month is posted on the schedule page. Readings and assignments may change. Deadlines will be announced in class too and are also on Carmen.

This is a 3-credit course, graded on a letter-grade (A, B, C, D, E) basis. Students are expected to attend class meetings, complete reading and assignments, as well as actively participate in class discussions.

Homeworks (75%): Three homework assignments will be due by the beginning of class. They will be turned in through Carmen. No late homeworks will be accepted.

Participation (10%): Attendance in class is strongly recommended. Active participation is required, since it will help you get the materials. Due to the course being online this year, if you cannot connect to the Zoom meeting, you are required to post on Carmen what you did on the day you missed the Zoom meeting and whether you have questions about the material.

Short answers (15%): Every week, you will reply to a short prompt, reflecting on what you have learned: what you have mastered or are still struggling with. The goal of these short answers is to evaluate your progress and make sure you are keeping up with the materials.

Grades will be assigned using the standard OSU scale.

The course usually takes place in our Computational Linguistics lab, where we are making sure that all software and packages needed for the course are available. There are detailed notes on Carmen to help you download and install the software on your computer. Email me if you are having issues with any of this *prior* to May 11!

Materials for each unit will be posted on the website and on Carmen, as will the slides presented in class. Datasets which cannot be made publicly available will be on Carmen. Assignments will need to be turned in through Carmen.

Note that email from Carmen is sent to your official email address (Name.Number@osu.edu). You should read email sent to your official OSU account on a daily basis.

If you know you won't be able to make a deadline, talk to me before you miss the deadline! Once the deadline is passed, barring a very good reason, no late homework will be accepted.

It is the responsibility of the Committee on Academic Misconduct to investigate or establish procedures for the investigation of all reported cases of student academic misconduct. The term “academic misconduct” includes all forms of student academic misconduct wherever committed; illustrated by, but not limited to, cases of plagiarism and dishonest practices in connection with examinations. Instructors shall report all instances of alleged academic misconduct to the committee (Faculty Rule 3335-5-487). For additional information, see the Code of Student Conduct (link to the Code).